WLAN Service Coverage Based on PixelFlow Predictions

Increase of WLAN network deployments lead to the need of developing tools to predict coverage in terms of available service. In this paper we propose to establish service coverage based only on approximate floor plans by using the so-called PixelFlow algorithm. This algorithm is based on discrete version of the Huygens principle and appears to be rather robust to approximation in the floor plan. Measurements of service performances have been undertaken and used to calibrate the prediction results. Since the conventional calibration based on prediction errors is not the goal of service coverage prediction, a new metric has been developed to quantify differences between predicted and measured boundaries of the service coverage. Despite the complex impact of inaccuracies related to floor plan, wall material, radio propagation and WLAN protocols, it appears that the service coverage prediction proposed here is suitable to ease the radio network design of indoor Wi-Fi system.